[genoPlotR-help] genoplotR

[Lionel Guy]

Hi Ronnie!

Thanks for your interest in genoPlotR! See my comments inline.

> On 20 Nov 2014, at 17:02 , Ronnie De Jonge <rojon at psb.vib-ugent.be> wrote:
> 
> I've however run into a rather strange error, perhaps bug. 
> 
> I have two dna_segs for which I have blastn alignment (local bl2seq with -D 1). Whenever I trim the alignment file (in fact pre-trim by setting a rather stringent e-value) to obtain only two 'major' alignments (one of ~50 Kbp and another, directly adjacent of ~11 Kbp) and use the plot_gene_map function to plot this comparison (that I loaded using try(read_comparison_from_blast("blastfile") ) it only wants to plot one of the two alignments. If I were to add a third line with 12 * 0\t to the original blast file and use the same methodology to import it will in fact plot this second alignment. 

This is indeed a bit weird. Since the file is read with the standard read.table function from R, it would be surprising if it was a real bug. The easiest explanation is that your file lacks a EOF character at the end, and that the last line is ignored. What if you just put a newline at the end of the last line? Would that work? Otherwise, could you send me the file and your code so that I can see whether I can reproduce the error? 

> Considering this is a very easy example (where I was able to spot the problem), and I'm interested in plotting more complex (more divergent species, consequently more alignments) comparisons, I obviously would like to be sure that what's plotted actually makes sense...
> 
> Another question I have relates to the color gradient. I like the colors, but I would like to know what they represent (kind of legend figure would be nice). Probably I missed it in the vignette but couldn't find it. 

It's somewhere on my todo list, but never had time to implement it. You can have a look in the documentation of apply_color_scheme for more info. By default, it takes the direction of the hit (direct or complementary) to attribute red/blue. For the hue, it looks at the range of percent id in all comparisons file, and applies a linear scale (brightest = strongest).

> Final question is a bit of personal / file-specific comment. Whenever genbank files are imported for which the gene coordinates are surrounded by '<' and '>', genoPlotR can't import these gene structures. Although the official format suggests indeed these shouldn't be there; many (also NCBI-derived; see e.g. here: http://www.ncbi.nlm.nih.gov/nuccore/45477378?report=genbank) do have these. I simply use a sed '/gene/ s/[<>]//g' command to remove them; but some other users might not be so familiar with such command.

I know. This is far from perfect, and a better genbank file parser is also on my todo list somewhere. Haven't had time to play with it though... But any contribution is welcome ;)

Good luck with genoPlotR! Don't hesitate to come back to me if you have further questions/problems!

Cheers,

Lionel

> Looking forward to your response.
> 
> With kind regards,
> Ronnie de Jonge
> 
> 
> -- 
> ==================================================================
> Ronnie de Jonge, Ph.D. 
> 
> Tel:+32 (0)9 33 13537  fax:+32 (0)9 33 13809
> Mobile: +32 (0) 468 14 51 29
> VIB Department of Plant Systems Biology, Ghent University
> Bioinformatics and Evolutionary Genomics Division, BEG
> Technologiepark 927, B-9052 Gent, Belgium
> Email: ronnie.dejonge at psb.ugent.be     
> Website: http://bioinformatics.psb.ugent.be
> ==================================================================
>